The reason I am choosing Net A Porter and Farfetch is because I am interested in fashion. And from where I come from, there is no such online fashion retailer like these two, so I find this unusual and fancy. Another reason is that I would like to work in industries like online fashion retailers doing data analysis work.
Those two online fashion retailers are all from British and basically they sell the same product, so I am wondering when people are considering buying clothes online, what is their altitude towards these two online fashion retailer.
Therefore, for my analysis, I will be focusing on text analysis which will show people’s altitude toward these two shops and maps to show from where are people talking about each shop. And I will be extracting their official twitter account and do some basic analysis on it. Overall, there will be four main analtsis which are:map, sentiments, wordcloud, timelne.
To start analyzing the data, the first thing I should be doing is to get twitter data through API. For the map part, I used R package called streamR, and for sentiment, word cloud and timeline parts I used R package called twitterR.(Code will be found on my github website:https://github.com/YingLi99/615-twitter-mining).
For this part, I will be talking about the analysis I did. The analysis mainly consists of different plots and maps for the two online fashion retailers.
Based on the map for Net A Porter, we can see that the people talking about net a porter are most from Los Angeles, San Francisco and New York. It makes sense because from my perspective of view, those three places are the place of fashion, that is to say, most of the fashion bloggers are from those three places.
The situation is the same for Farfetch, however comparing two maps, we can clearly see that the map for Net A Porter has more points than the map for Farfetch.
I think this may be the reason that Net A Porter is more popular in America than Farfetch.
From the word cloud for Net A Porter, we can see that the words with the highest frequency are all positive words, same situation is applied for the word cloud of Farfetch. However, we can only see the words frequency of these two online fashion retailers, it is hard for me to compare people’s attitude towards these two shops.
To really get hold of the people’s feelings towards these two shops. Doing a sentiment analysis is necessary. Therefore for the next part of my analysis I did a sentiment analysis.
## Selecting by n
## Selecting by n
To look at those two plots separately, for Net A Porter, we can see that most of the feeling are positive and there are only a little words that are negative.
For Farfetch, the situation is quite different, the number of positive words are relatively low. And there is an unexpected negative words with high frequency.
To compare those two plots, it is quite obvious that people have a better feeling towards Net A Porter than towards Farfetch.
For now, from the map analysis and sentiment analysis, we can conclude that Net A Porter might be better than Farfetch in terms of people’s feelings. To confirm this, I will collect data from the official twitter account of these two shops.
From these two plots, we can see that in fact the favorites and retweets amount of Net A Porter and Farfetch twitter is pretty low, my guess is that twitter is not that popular these days while instagram is.
From these two plots, we can easily see that there is a high peak for both of the plots. I check the twitter and find out the post with the highest favorite amount, my guess is that this post has information that most people agree with.
Overall, to find people’s attitude towards Net A Porter and Farfetch, I did several analysis and find out that people talked more about Net A Porter and have a higher positive feelings about it. However, the total twitter ammount for Net A Porter and Farfetch is different, this might be the reason why there is difference in people’s attitude towards them.